search

zsef123 / EfficientNets-PyTorch

A PyTorch implementation of " EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks."
308 stars 47 forks source link

the meaning of depth div #8

Open twmht opened 5 years ago

twmht commented 5 years ago

Hi

what is the meaning of depth_div used in new_ch?

https://github.com/zsef123/EfficientNets-PyTorch/blob/master/models/effnet.py#L81

wuhy08 commented 5 years ago

It seems to me that we want the # of channels in each MBconv block to always be the multiple of depth_div.

For example, in the implementation, depth_div 8. This means # of channels should be 8, 16, 24, 32, 40, etc.